The dentate gyrus (DG) of the hippocampus plays a key role in memory formation by transforming patterns of cortical inputs into new patterns of output to the CA3 area. Although the cellular and synaptic basis of this important transformation remain poorly understood, two excitatory cell types in the DG, granule cells (GC) and hilar mossy cells (MC), play a major role. MCs mediate an intrinsic, hetero-associative (GC-MC-GC) excitatory loop, receiving powerful input from a relatively small number of GCs, and providing highly distributed excitatory output to a large number of GCs. MCs project their associational and commissural axons to the ipsi- and contralateral inner molecular layer of the DG, where they synapse onto proximal dendrites of GCs. Moreover, MCs also project their axons along the septotemporal axis of the hippocampus, thereby connecting functionally diverse areas of this structure. By projecting to most areas of the DG along the septotemporal axis, MCs could provide important contextual content to the information arising from the cortex. In order to understand how information is processed in the DG and how dysregulation of this circuit may contribute to disease, a better knowledge of the hetero-associative GC-MC-GC circuit and its dynamic properties is required. We have recently reported that MC-GC synapses undergo a novel presynaptic, NMDA-receptor independent form of long-term potentiation (LTP) that requires postsynaptic brain-derived neurotrophic factor (BDNF)/TrkB and presynaptic cyclic AMP(cAMP)/PKA signaling. We hypothesize that this novel form of plasticity enhances GC output at the associative MC-GC recurrent circuit, and may contribute to DG-dependent forms of learning and brain disease, such as epilepsy. A large number of questions regarding this circuit remain unanswered. Preliminary data indicates that MC-GC LTP is induced in vivo by experience and epileptic activity, is critically regulated by endogenous systems (e.g. endocannabinoid and adenosine signaling), and it can be accompanied by LTP of inhibitory transmission. Here, using a combination of experimental approaches both in vitro and in vivo, we aim to (1) characterize the synaptic learning rules of MC plasticity, (2) identify the molecular mechanism underlying MC-GC LTP, (3) determine the properties and mechanism underlying inhibitory LTP, and (4) determine the functional relevance of MC plasticity in vivo. By identifying the main properties and mechanisms of activity-dependent plasticity in a crucial recurrent circuit in the DG, our proposed studies may not only improve our understanding of the precise role of this circuit in DG information processing and memory encoding, but also assess how dysregulation of this circuit may contribute to brain disease, including epilepsy, anxiety, schizophrenia and depression.